Evaluation of Cardiac Output Estimation with Veno-Venous Extracorporeal membrane Oxygenation Support: in vitro study

Veno-venous extracorporeal membrane oxygenation (V-V ECMO) is used to treat refractory lung failure, a condition that often results in right ventricular dysfunction, which can exacerbate both morbidity and mortality. Monitoring cardiac output as an indication of ventricular function for V-V ECMO may assist in diagnosing patients before significant deterioration. It also provides the benefit of monitoring V-V ECMO flow to ensure proper oxygenation of tissues. However, no gold-standard technique exists to measure cardiac output with V-V ECMO. Cardiac output algorithms based on thermodilution or pulse contour analysis from arterial pressure waveforms have been routinely used in the intensive care unit; however, these have not yet reached clinical acceptance for patients on V-V ECMO due to the few clinical studies, which are limited to variability, gold-standard reference cardiac output measurement and investigated range of conditions. This study therefore investigated the performance of estimated cardiac output from pulse contour analysis in a repeatable bench top mock circulation loop where a wide range of cardiovascular conditions could be investigated, and cardiac output could be directly measured. The wide range of cardiovascular conditions included changes in right ventricular contractility, pulmonary vascular resistance, V-V ECMO flow, and aortic compliance. Three different pulse contour analysis algorithms (Windkessel model, Windkessel with Resistive Capacitive (RC) delay and Liljestrand-Zander model) were investigated and chosen based on greater performance of intensive care unit patients. This in vitro study demonstrated that the Windkessel model and Windkessel with RC delay had strong correlation (R2: 0.79 and 0.82 respectively) with no bias and low errors (0.00 ± 0.29 and 0.00 ± 0.26 respectively), whilst the Liljestrand-Zander model had weak correlation (R2: 0.32) and greater error (0.00 ± 0.48). This study therefore suggests that the Windkessel model and Windkessel with RC delay model could be used to continuously monitor cardiac output for patients support with V-V ECMO. Future work is to verify these results in animal and clinical studies.Clinical Relevance- The clinical utility of continuously monitoring cardiac output (from pulse contour analysis of arterial pressure signal) may assist clinicians in earlier detection of heart dysfunction for patients with V-V ECMO. Consequently, clinicians may more rapidly initiate further investigations or administer medical therapies to reduce morbidity and mortality.